When individuals read or hear language, they must construct a syntactic parse that connects the various words in a sentence together and allows the individual to comprehend the message. However, this process may not be as straightforward as intuition might imply. Self-organizing models of sentence processing claim that the mind spends significant amounts of time in states of partial aggregation, forming partially coherent parses that do not from a cohesive whole. This framework also predicts that interference from partially coherent syntactic structures plays a role in the comprehension difficulties of poor readers at the level of the sentence. The aim of the current research project is to examine the role of partial coherence in human sentence processing, and to investigate the role of partially coherent structures as a source of difficulty in poor readers. The current project uses the technique of Visual World eye tracking, in which participants hear language over headphones while looking at a visual scene. By measuring trends in the distributions of looks, we can figure out the process by which people make sense of the language in the context of the visual scene, and we can discover ways in which participants fail to reach states of complete coherence. We make systematic comparisons between stimuli presented in the spoken form versus the written form to understand how reading specifically may go awry. In certain designs, we measure reading times on individual words using Self-Paced reading, where readers see the words one at a time and push a button to move to the next word. In other designs, we first have participants read sentences, and then perform tasks involving visual scenes that test their comprehension of the language. The current project studies reading difficulty in individuals with a wide range of abilities, including young adults for whom difficulties with reading have been a hindrance to educational success. We believe that people plateau in states of partial reading success because the mind can fall into a state of gridlock. This project probes the circumstances under which such gridlock arises in the domain of reading as a step toward understanding individual and societal flexibility/rigidity generally.